Identifying the neural substrates of schizophrenia is a key step in developing more targeted treatments or identifying molecular mechanisms. The tools of neuroimaging have substantially facilitated this search. For approximately the past decade, we have been exploring the hypothesis that patients with schizophrenia suffer from a "misconnection syndrome": a disturbance in a widely distributed network of brain regions. This misconnection leads in turn to impairment in the ability to integrate information ("thought disorder") that is the hallmark of schizophrenia. The proposed project examines both structural connectivity (using diffusion tensor imaging, or DTI) and functional connectivity using functional magnetic resonance imaging (fMR) to examine the "default network" during random episodic silent thought ("REST"). We will employ both a cross-sectional and a longitudinal design in order to examine whether impairments in structural and functional connectivity progress over time, based on our preliminary work suggesting that this is the case. We will collect MR data at baseline in first episode patients, chronic patients, and healthy normal volunteers. A subset of the first episode patients will be studied when they are still neuroleptic naive. Follow-up MR data will be obtained at two time points. Only a few studies to date have examined first episode patients. Ours will be one of the first to compare first episode patients and chronic patients, using a matched control group. There are very few studies examining the issue of progression or of medication effects with DTI. Neuroleptic naive subjects have never been examined, nor has DTI been applied in a longitudinal design. DTI and fMR provide different and complementary data about the brain;additional power is provided if this information can be integrated in a single analysis. We will use novel statistical techniques for "data fusion" to integrate the information from DTI measures of structural connectivity with fMR measures of functional connectivity. Thus the proposed work with is highly innovative.